Most sales look good on the revenue line.
Then the margin is gone, customers learn to wait for the next one, and the stock that really needed to move is still on the shelf. Mercantile is what I'm building for Shopify stores: the financial analyst a small store never hired, deciding which offer to run, or whether to run one at all.
Sale week
Revenue
Profit
Offers are made blind, or rented.
Most stores design their offers one of two ways. They guess: twenty percent off, because a competitor did it, or because it's a holiday. Or they hire someone, a financial analyst or a cost controller, to study the numbers and design offers that sound great without losing money.
Most small stores can't afford the second, so they do the first. Mercantile does the second job for them. You install it in your store, connect it, and it reads what you already have: your products, your stock, your orders, your costs, and how past offers went.
Guess
Cheap, and blind.
Hire an analyst
Careful, and expensive.
Mercantile
The careful work, installed.
It starts from where the store is under strain.
Most offer tools start from demand: what will people buy? Mercantile starts from pressure: where is the store under strain? Stock that isn't selling. Items close to their expiry date. Products with enough margin to pay for a promotion. Products that already sell well at full price and should never be discounted. Customers who used to buy and stopped.
Some pressure you want to release. Some you want to keep.
- Not selling
- Close to expiry
- Room in the margin
- Sells well at full price: protect it
- Bought by customers who stopped coming
A handful of careful moves.
From that map, Mercantile puts together a few possible plays, each with different economics. A discount that's fenced in so it can't spread. A bundle that moves slow stock alongside something people already want. A threshold that raises the size of an order. A win-back offer for customers who went quiet. A staged release for aging stock. A short window for something that's suddenly trending.
- 1
More visibility
Show it where people look.
- 2
A gentle bundle
Pair it with something they want.
- 3
A markdown
Lower the price, carefully.
- 4
Clearance, only at the end
Clear what's left.
For each one, it projects what matters. Profit, not only revenue. How much margin is exposed. Who would have bought anyway. Whether it teaches customers to wait for sales.
Reached by the offer
Would have paid full price
Margin given away for nothing
What a week's advice might look like.
Mercantile · This week
Don't discount the serum. It already sells at full price, and a discount would only give margin away.
Use it as the anchor in a three-item bundle with the two cleansers that aren't moving.
The sunscreen batch expires in ninety days. Give it more visibility this week. If it hasn't moved in a month, add it to a gentle bundle.
If the bundle hasn't sold after two weeks, switch to a free gift instead of a deeper discount.
The cleansers: carried by it
The serum: never discounted
Sometimes the answer is no sale.
Don't run a sale this week.
Some weeks, that's the right advice. The problem isn't the price. It's that people can't find what they're looking for, or the product page doesn't explain it. A tool that always recommends a discount is a tool that profits from discounts. I want one that tells the store the truth.
This week's pressure
Fenced discount
Bundle
Threshold
Win-back
Staged release
Trend window
No sale this week
It recommends, or it acts. You choose.
Mercantile can recommend a play and wait for your approval, or run it on your behalf inside limits you set: a margin it can never go below, products it may never touch, a limit on how many offers run at once, and a rule for when to stop an offer that isn't working. Every offer it runs comes with its own way to stop. It can also tell you what to retire: the products that cost more to keep than they bring in.
Every play waits for your approval.
Margin floor
Protected products
Offers at once
Stop if it isn't working
Tripped: the offer stops on its own.
The offer runs
Margin floor
Protected products
Offers at once
Stop if it isn't working
Tripped: the offer stops on its own.
The offer runs
And it learns what works in your store.
After each offer, it looks at what actually happened, and learns which kinds of offers work under which conditions, in your store specifically. It starts with Shopify, because that's where so many small stores already live. Later, stores on other platforms can connect through their own integration.
| Row | Play | Relieves | Margin at risk | Ends when |
|---|---|---|---|---|
| 1 | Serum bundle | Slow cleansers | Low | Two weeks, or when the cleansers sell |
| 2 | Sunscreen visibility | Expiring batch | None | When the batch is gone |
| 3 | Win-back | Lapsed customers | Medium | After one message |
| 4 | No sale | Nothing to release | None | Next week |
Serum bundle
- Slow cleansers
- Low
- Two weeks, or when the cleansers sell
Sunscreen visibility
- Expiring batch
- None
- When the batch is gone
Win-back
- Lapsed customers
- Medium
- After one message
No sale
- Nothing to release
- None
- Next week